← All incidents

[Eval][v1][zh] Third-party payment gateway timeouts cascade into checkout outage

service: checkout-svccreated: 6/2/2026, 11:56:14 PM

Raw incident context

Time: 18:40 UTC. checkout-svc starts returning HTTP 504 to ~78% of checkout attempts.

Symptoms:
- p99 latency: 28s (hitting our 30s gateway timeout)
- Success rate: 22% (down from 99.5%)
- Thread pool utilization: 100% (all 200 worker threads blocked)
- Inbound queue depth: 4200 requests (queue limit 5000)

Downstream calls (from APM):
- Stripe Connect API (/v1/payment_intents): p99 jumped from 800ms to 27s
- Stripe status page: "Investigating elevated latency for Connect endpoints in us-east-1" (posted 18:35 UTC)
- All other downstream deps (auth-svc, fraud-svc) normal

Our config:
- Stripe call timeout: 30s (matches inbound)
- No circuit breaker on Stripe client
- No bulkhead — Stripe calls share the main worker thread pool

Customer impact:
- Failed checkouts: ~3000 in 5min
- Estimated lost GMV: $180k (so far)

On-call:
- 18:38 — pager
- 18:41 — confirmed Stripe is the culprit (status page + APM)
- 18:43 — debating: wait it out vs disable Stripe path entirely

Summary

SEV1

checkout-svc 因依赖的 Stripe Connect API 延迟飙升(p99 从 800ms 升至 27s)导致线程池耗尽,78% 的结账请求超时失败,影响约 3000 笔交易,损失约 $180k GMV。

Severity reasoning: 用户可见的结账失败率达 78%,直接导致收入损失($180k/5min),且影响范围广(所有使用 Stripe 支付的用户),属于 SEV1。

deepseek-chat·prompt v1·output: zh·10106ms·1781↑ / 1616↓ tok·$0.00226

Root cause hypotheses

  • highStripe Connect API 在 us-east-1 区域出现延迟问题,导致 checkout-svc 的 HTTP 调用超时。

    Evidence: Stripe 状态页面于 18:35 UTC 发布 'Investigating elevated latency for Connect endpoints in us-east-1';APM 显示 Stripe /v1/payment_intents 的 p99 延迟从 800ms 升至 27s。

  • highcheckout-svc 未配置断路器或舱壁隔离,Stripe 调用阻塞了所有工作线程。

    Evidence: 线程池利用率 100%(200 个线程全部阻塞),入站队列深度 4200(接近限制 5000),且 Stripe 调用超时设为 30s 与入站超时一致,无回退机制。

  • mediumStripe 客户端连接池耗尽或 DNS 解析缓慢,加剧了延迟。

    Evidence: 无直接证据,但线程阻塞时连接池可能被占满;可检查连接池指标。

Investigation checklist

  1. 确认 Stripe 状态页面最新更新
    curl -s https://status.stripe.com/api/v2/incidents/current.json | jq '.incident.status'

    Expected: 返回 'investigating' 或 'identified',确认 Stripe 仍在处理中

  2. 检查 checkout-svc 线程转储,确认阻塞线程
    kubectl exec deploy/checkout-svc -- jstack $(pgrep -f checkout-svc) | grep -A 20 'BLOCKED' | head -100

    Expected: 大量线程处于 BLOCKED 状态,等待 Stripe HTTP 调用完成

  3. 查看 Stripe 客户端连接池指标
    curl -s http://checkout-svc:8080/metrics | grep -E 'stripe_connections|http_client_active'

    Expected: active_connections 接近最大值,或连接池耗尽

  4. 检查入站队列深度和超时配置
    kubectl exec deploy/checkout-svc -- curl -s localhost:8080/actuator/health | jq '.details.queueDepth'

    Expected: 队列深度接近 5000 限制,确认背压

  5. 验证其他下游服务是否正常
    kubectl exec deploy/checkout-svc -- curl -s -o /dev/null -w '%{http_code}' http://auth-svc:8080/health

    Expected: 返回 200,确认 auth-svc 正常

Mitigation plan

  • 立即启用 Stripe 客户端的断路器,设置失败阈值(例如 5 秒内 50% 请求失败则断开),并配置快速失败回退(返回友好错误或重试队列)。

    Risk: 断路器误触发可能导致正常流量也被拒绝;但当前 Stripe 已降级,风险可控。

    Rollback: 回滚配置:将断路器阈值调高或禁用断路器,恢复原始配置。

  • 临时将 Stripe 调用超时从 30s 降低到 5s,减少线程阻塞时间。

    Risk: 部分正常慢请求可能被提前超时,但当前 Stripe 延迟极高,5s 超时可快速释放线程。

    Rollback: 恢复超时配置为 30s。

  • 如果 Stripe 问题持续,考虑将支付降级为仅支持非 Stripe 支付方式(如 PayPal),或在前端隐藏 Stripe 选项。

    Risk: 功能降级影响用户体验,但可保住部分收入。

    Rollback: 重新启用 Stripe 支付选项。

Customer impact

约 78% 的结账尝试失败,影响约 3000 笔交易,预计损失 $180k GMV。用户看到 HTTP 504 错误,无法完成购买。当前无 ETA,等待 Stripe 修复。

Postmortem draft

事后复盘

摘要

[2-3 句概述]

时间线

  • 18:35 UTC: Stripe 状态页面报告延迟问题
  • 18:40 UTC: checkout-svc 开始返回 504,告警触发
  • 18:41 UTC: 确认 Stripe 为根因
  • [添加缓解时间]

影响

  • 失败结账数: ~3000
  • 损失 GMV: ~$180k
  • 持续时间: [待补充]

根因

Stripe Connect API 延迟飙升导致线程池耗尽。

做得好的

  • 快速识别 Stripe 问题
  • 状态页面确认

做得不好的

  • 无断路器或舱壁
  • 超时配置未与下游匹配

行动项

  • [ ] 为 Stripe 客户端添加断路器
  • [ ] 实现舱壁隔离
  • [ ] 调整超时策略
  • [ ] 添加降级支付路径

Follow-ups

  • P0为 Stripe 客户端添加断路器(如 Hystrix 或 Resilience4j)checkout-svc 负责人
  • P0实现线程池舱壁隔离,将 Stripe 调用分离到独立线程池checkout-svc 负责人
  • P1调整 Stripe 调用超时策略,根据 SLA 设置更短超时并加入重试checkout-svc 负责人
  • P1添加支付降级路径,当 Stripe 不可用时自动切换备用支付平台团队
  • P2监控 Stripe 状态页面并集成告警on-call SRE